Introducing: Reprap Ultrascale (Idea)

I have an idea how to build an extremely large area 3d printer and would like to release it into the wild as I don't have time to build it. After thinking about it for about a week, I'm sure it would work, but I have too many other projects to work on.

The basic idea is quite simple: imagine a 2d robot arm working in the XY plane. It doesn't matter really what type it is (SCARA, parallel SCARA, whatever). Attached to the bottom of the end effector would be a plastic "puck" that would slide across a glass pane. On the top of the end effector would be an extruder that would point *up* instead of down. For the Z axis, a print bed would be raised and lowered on top of the nozzle.

Because the arms don't have to support their own weight, horizontal bending and flexing is completely eliminated. This allows us to scale up the arms almost indefinitely, assuming we have a flat surface to guide them smoothly.


For those curious, printing upside down is definitely doable.
[www.youtube.com]

Edited 1 time(s). Last edit at 03/28/2015 03:05PM by 3Derping.

... you have 'eliminated' the weight issues of the arm, but not the weight of the printed object ... and don't forget all the warping related issues, which are related to size ... and now will be too connected to gravity ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I'm pretty sure gravity warps parts to the same degree either right side up or upside down.

Also, large print warping isn't a showstopper. There have been quite a few large scale printers that seem to manage just fine.

Now I will admit that printing this way will need greater surface area on the "print ceiling" (to keep the print from falling off due to weight) but that too can be managed via support structure (raft + stringers that connect from overhangs to the print ceiling). To deal with the raw weight of the print, you could counter balance some of the load. It's not like the print needs to accelerate much.

Is there anything else that bothers you about this idea?

Edit: After thinking about it some more, yes, I see how being upside down would hurt layer lamination rather than help it.

Edited 4 time(s). Last edit at 03/28/2015 08:00PM by 3Derping.

I guess this begs the question WHY?

Why would this design be better than more classic designs?


What about a glass plate on top, or teflon coated plate.
A magnet on the top of extruder and a magnet on plate above extruder magnet
on top of plate to carry the weight?
Maybe some ball bearings on bottom of top magnet to make it slide around easily

I suppose one could use a 10ft by 10ft piece of plate window glass (Software compensation for sag)
to print out sections of your new house?


How about just driving top magnet with core XY rig
Magnet on top of extruder just following around?

Rare earth magnets could hold up 100 pounds of extruder?

Instead of mechanical drive -- Perhaps little thrusters to move the puck around
which would move the extruder around

Sky is the limit?
but is it a GOOD design?

Yeah, I guess you're right: printing car sized objects is just not worth the effort. /s

I would agree that a mounted extruder is best *if* we could find a suitable bearing for it to glide easily. The problem is this: if you used ball bearings, how would they recirculate in 2 directions? Also, if we use solid plastic based bearings, how will we lower the coefficient of friction enough for them to glide under load? Not only does the top magnet have to slide but also (and more importantly) the bottom magnet has to follow without hesitation.

The problem with corexy is that you need linear guides if you wish to make movements accurately. Sure, you can use string like this, but you'll be flexing the string when you accelerate.

And yes, I do think it is a good design, for the reasons stated earlier. Why do you think it would be a bad design? Just saying something is bad does not make it bad. I don't mind you making fun by suggesting thrusters, but please try and be constructive.

Edited 1 time(s). Last edit at 03/28/2015 11:03PM by 3Derping.

I think another big advantage to putting the extruder on the bottom is added rigidity. One could pour some cement, line the top of it with a glass plane, and then bolt the arm mountings into the cement while resting the arms on the glass. It doesn't get much more rigid than that.

Or you could just bolt everything to your garage floor.

Edited 1 time(s). Last edit at 03/28/2015 11:45PM by 3Derping.

Ever heard of air hockey? Hovercraft? No friction with air bearings.

It's definitely an interesting idea. Do you have a target dimension? If so you could get a rough idea of the mass of the object being printed, then do some experiments to find out if the build plate adhesion will be strong enough to hold it and go from there.

---

[scara3dprinter.wordpress.com]

What if the build plate was lined with an array of holes? One could pause the print midway (where it was starting to get too heavy) and screw through the buildplate and into the print. Then weight shouldn't become a problem until the layers start to delaminate.

This kinda reminds me of this guy: [www.rooiejoris.nl]

Edit: @ see3d If you think that would work, then awesome. I didn't know they had the capacity to hold that much weight. Would the air cushion stay constant enough to not mess up the Z position?

Edited 2 time(s). Last edit at 03/29/2015 12:42AM by 3Derping.

That could work if you didn't mind drilling little holes in your print, but what about a vacuum table similar in concept to what machinists use, there couldn't be lots of grooves for obvious reasons, but a number of small holes pulling a vacuum could work.

---

[scara3dprinter.wordpress.com]

Something like this. Just tape over the holes not needed to maintain the vacuum.


View from below


Pull a vacuum from the hole in the bottom plate.

---

[scara3dprinter.wordpress.com]

Looks great! With this, you could probably lift a lot of weight. Would the cost be high?

Hard to say, it crossed my mind that I could use this on my SCARA printer and be able to print ABS without a heated bed, but it is only 9 inches square. The aluminum for that would only cost around $20. A brake bleeder hand pump could be used to pull a vacuum.

---

[scara3dprinter.wordpress.com]

The way an air bearing works is that air is pumped between two very flat surfaces. The air pressure is reduced because it has to speed up to get out of the small space. This means the the air gap is balanced between the 14 PSI air pressure and just enough gap to get the air out. If air is pumped into the puck with a flexible tube, then it should work (with proper engineering and very flat surfaces). The challenge will be how to feed the material in to 3D print from the side. My gut tells me that it can work, but is not the best approach.

... for the 'gliders' - look at this images: [www.tretter.de]

This are common 'omnidirectional' ball-gliders ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

As far as cost, I'm assuming an air bearing could be made, yes? So that wouldn't be to bad. How's the cost for a compressor that can maintain the right psi?

All of my knowledge of pneumatics fits in a thimble, so if you say it would work, I'll take your word for it . Personally, I'd just put the puck on the bottom surface and print onto the top. That way, a low-tech teflon pad could be used.

I think feeding the filament in is another big issue with using a guide plate, in any orientation. What's the minimum bend radius of 1.75 filament? I think it's around 4 inches (100mm). This would really add to the length of the end effector. Once we have a puck, a 4 inch space for the filament to turn and a nozzle, we're looking at a 6-8inch long print head.

Edited 1 time(s). Last edit at 03/29/2015 05:49PM by 3Derping.

For what its worth, I've built hot ends where the filament came in perpendicular to the nozzle, so the print head could be less than 1 inch long

---

[scara3dprinter.wordpress.com]

FWIW
They move around large heavy spacecraft (16000 lb spacecraft plus 25000lb holding fixture)
on air tugs ---- four 4 x 4 ft hover craft (air bearings) on a painted concrete floor
can be pushed around by hand --- but once it gets going -- it keeps going
-- so stopping more important than starting --- bleeding air out acts to brake
but sudden stop could ruin a $500,000 spacecraft.

200 - 3000 psi air feed

Craft gets moved from assembly area, to test area, to thermal vacuum chamber and back.

other precision machinery moved same way

Air bearings do sound cool. Perhaps they would be needed on extremely large builds.

I was also thinking about the lowest cost type of 2d positioning system. Perhaps a V plotter type system? [2e5.com] It would have to be held tight by a third string with a weight or spring on it.